Desalination and Reuse of High-Salinity Shale Gas Produced

Desalination and Reuse of High-Salinity Shale Gas Produced Water:Drivers, Technologies, and Future DirectionsDevin L. Shaffer, Laura H. Arias Chavez, Moshe Ben-Sasson, Santiago Romero-Vargas Castrillon,Ngai Yin Yip, and Menachem Elimelech*

Department of Chemical and Environmental Engineering, Yale University, P.O. Box 208286, New Haven, Connecticut 06520-8286,United States

ABSTRACT: In the rapidly developing shale gas industry, managing produced water is a major challenge for maintaining theprofitability of shale gas extraction while protecting public health and the environment. We review the current state of practice forproduced water management across the United States and discuss the interrelated regulatory, infrastructure, and economicdrivers for produced water reuse. Within this framework, we examine the Marcellus shale play, a region in the eastern UnitedStates where produced water is currently reused without desalination. In the Marcellus region, and in other shale plays worldwidewith similar constraints, contraction of current reuse opportunities within the shale gas industry and growing restrictions onproduced water disposal will provide strong incentives for produced water desalination for reuse outside the industry. The mostchallenging scenarios for the selection of desalination for reuse over other management strategies will be those involving high-salinity produced water, which must be desalinated with thermal separation processes. We explore desalination technologies fortreatment of high-salinity shale gas produced water, and we critically review mechanical vapor compression (MVC), membranedistillation (MD), and forward osmosis (FO) as the technologies best suited for desalination of high-salinity produced water forreuse outside the shale gas industry. The advantages and challenges of applying MVC, MD, and FO technologies to producedwater desalination are discussed, and directions for future research and development are identified. We find that desalination forreuse of produced water is technically feasible and can be economically relevant. However, because produced water managementis primarily an economic decision, expanding desalination for reuse is dependent on process and material improvements toreduce capital and operating costs.

■ INTRODUCTION

The application of horizontal drilling and hydraulic fracturingtechniques has unlocked a vast natural gas resource from shaleformations. With the addition of shale gas reserves, estimatedglobal technically recoverable gas reserves have increased byover 40% since 2010,1 prompting the International EnergyAgency to speculate about a “Golden Age of Gas” that will becharacterized by high energy demand in urbanizing regions andlow-cost, widely available shale gas resources.2 In the UnitedStates, where shale gas production techniques were developed,shale gas is recognized as a “game changer” in the natural gasmarket,3 and the U.S. is projected to move from a net naturalgas importer to become a natural gas exporter.3

The rapid development of shale gas in the United States hasbeen controversial. Proponents advocate for shale gasproduction as an economic boon4 and a potential bridge to a

low-carbon future.5 Opponents contend that the potential fordrinking water contamination from shale gas wells,6,7 negativeenvironmental impacts of produced water management,8 andpotential for leakage of methane,9 a potent greenhouse gas,outweigh any benefits. Some states in the United States10,11 andother countries2 maintain effective moratoriums on drilling anddevelopment of shale gas wells.Managing the produced water that is associated with shale

gas is one of the biggest challenges for preserving the favorableeconomics of shale gas development and protecting humanhealth and the environment.12,13 Growing restrictions on

Received: May 2, 2013Revised: July 18, 2013Accepted: July 25, 2013Published: July 25, 2013

Critical Review

pubs.acs.org/est

© 2013 American Chemical Society 9569 dx.doi.org/10.1021/es401966e | Environ. Sci. Technol. 2013, 47, 9569−9583

pubs.acs.org/est

produced water disposal and eventual contraction of reuseopportunities within the shale gas industry will ultimately movethe industry toward desalination of produced water, but high-salinity produced water is especially challenging and energy-intensive to treat. Technological advances in shale gas producedwater desalination, similar to the advances in hydraulicfracturing that facilitated the development of the shale gasindustry, must be developed to promote expanded reuseopportunities that require higher quality water.14 The Barnettshale region in Texas served as a testing ground for therefinement and commercialization of horizontal drilling andhydraulic fracturing techniques that are now being appliedworldwide.15 Similarly, the unique produced water character-istics, shale formation geology, infrastructure, and regulatoryenvironment of the Marcellus shale region in the easternUnited States are promoting innovation in produced waterreuse. The produced water management constraints of theMarcellus shale region will eventually drive Marcellus shale gasproducers, and those in similarly constrained regions, todesalinate produced water to broaden its potential for reuse.We critically review desalination technologies that may be

implemented at shale gas well sites or centralized treatmentfacilities to enable reuse of high-salinity produced water outsideof the shale gas industry. This “external reuse” requires muchhigher water quality than current “internal reuse” practices thatrecycle produced water within the shale gas industry forsubsequent hydraulic fracturing operations. Produced watermanagement in the Marcellus shale region is highlighted toillustrate the myriad considerations that influence producedwater treatment decisions and the anticipated future applicationof desalination technologies. Mechanical vapor compression,membrane distillation, and forward osmosis are reviewed as thethree desalination technologies best suited for use under theincentives and constraints represented by the Marcellus shaleregion. The relative energy requirements and potential energysources for the three technologies are assessed and compared.

■ PRODUCED WATERA SHALE GAS BYPRODUCT

Categorization and Components of Shale GasProduced Water. Produced water is water from undergroundformations that is brought to the surface during oil or gasproduction, and it is the largest volume waste stream associatedwith this production.16 Shale gas production is the extraction ofnatural gas from shale plays, which are hydrocarbon-rich, low-permeability shale formations within known hydrocarbon-producing basins. Shale gas is considered an unconventionalnatural gas resource because the shale serves as both the sourceand the reservoir for the gas. In contrast, conventional sourcesare relatively high-permeability formations that are capped withimpermeable barriers. Natural gas from sources beneath or

adjacent to the formation accumulates beneath theseimpermeable caps.17

The extraction of shale gas is facilitated by horizontal drillingand hydraulic fracturing techniques that are able to access largeareas of the shale formation and increase the shale permeabilityto enable gas to flow.17 Shale gas produced water consists ofboth formation water and flowback from hydraulic fracturing.Flowback returns to the surface during the initial weeks of wellpumping after hydraulic fracturing has occurred, and formationwater is extracted over the lifetime of the well.17 Figure 1illustrates the categorization of shale gas and the components ofshale gas produced water.The quantity and quality of produced water from shale gas

wells vary by shale gas play, by well location within a shale play,and over the lifetime of the well. The highest rate of flowbackoccurs immediately after well pumping begins and diminishesover time.12 Drilling and hydraulically fracturing a shale gas wellis estimated to require approximately 2−4 million gallons ofwater (48 000−95 000 barrels (bbl), 8000−15 000 m3).17

Historical data show that median water use for developing ashale gas well in Texas is 2.8−5.7 million gallons (67 000−140 000 bbl, 11 000−22 000 m3), depending on the specificshale play,18 and the median volume of water used forhydraulically fracturing a horizontal well in Oklahoma is 3.0million gallons (71 000 bbl, 11 000 m3).19 In the Marcellusshale region, drilling and hydraulically fracturing a horizontalwell requires an estimated 2−7 million gallons (48 000−170 000 bbl, 8000−27 000 m3).13 The percentage of this initialvolume that is returned to the surface as flowback is specific tothe well and has been estimated as 8−15%,14 10−40%,12 9−53%,13 and 30−70%.17 The remaining volume is consideredformation water, though differentiating flowback from for-mation water is difficult because fracturing fluids become moresimilar to formation water the longer they remain in contactwith the formation.17

Long-term produced water volumes from shale gas wells varyacross different shale plays and ultimately depend on thelifetime of the shale gas well. Shale gas produced watergeneration can be compared between shale plays using themetric of gallons of water produced per million cubic feet of gasproduced (gal/MMcf). One shale gas producer reports long-term produced water (which we assume to be formation water)estimates of 200−1000 gal/MMcf in the Eagle Ford, Haynes-ville, and Fayetteville shale plays; more than 1000 gal/MMcf inthe Barnett shale play; and only approximately 25−200 gal/MMcf in the Marcellus shale play.20 A recent analysis of thefirst 4 years of production from Marcellus shale gas wells inPennsylvania showed the average brine (formation water)production to be approximately 700 gal/MMcf.14

Figure 1. Schematic illustrating the categorization of shale gas as one type of unconventional natural gas resource and indicating the two componentsof shale gas produced water: flowback and formation water.

Environmental Science & Technology Critical Review

dx.doi.org/10.1021/es401966e | Environ. Sci. Technol. 2013, 47, 9569−95839570

Current Produced Water Management Practices inthe United States. Management of produced water isprimarily an economic decision21 that is directly influencedby the produced water volume and quality, state and federalregulations, available infrastructure, and characteristics of thespecific shale play. Figure 2 illustrates the considerations thataffect produced water treatment and disposal decisions. Thevast majority of produced water in the U.S. is injectedunderground via the federal Underground Injection Controlprogram, either to maintain pressure in active formations or fordisposal. A survey of oil and gas produced water managementpractices in the United States in 2007 found that more than98% of produced water from onshore oil and gas wells isinjected underground.16 Underground injection will remain animportant component of produced water management, evenwhen desalination is employed for external reuse, because it is adisposal mechanism for the concentrated brine stream resultingfrom any desalination process.A small fraction of produced water is discharged to surface

water, managed by evaporation ponds, or beneficially reusedoutside of the industry,16 for example in livestock watering orirrigation.22 These management practices are typicallyemployed only for high-quality produced water with relativelylow dissolved solids concentrations, such as in the Big Hornand Powder River Basins of Montana and Wyoming in theUnited States.22 Direct beneficial reuse of produced water inthe U.S. is limited by the Clean Water Act to livestock wateringor agricultural uses west of the 98th meridian.23

Reuse of produced water for shale gas well development(which we will term “internal reuse”) is a growing trend,especially where this reuse is economically advantageous, suchas in the Marcellus shale region. Successful internal reuse canreduce the fresh water demands for subsequent hydraulicfracturing operations as well as reduce produced water disposalcosts. Internal reuse has expanded as shale gas producers haveexperimented with reusing produced water that has not beendesalinated.21 For internal reuse, produced water is oftenblended with fresh water to reduce the high dissolved solidsconcentration and mitigate its effects on fluid viscosity.20

Internal reuse of produced water with elevated concentrations

of dissolved solids must also consider factors such as corrosionof well materials,18 scaling that impedes gas flow to the well,14

and the effects of varying salinity on clay swelling within theformation.14

Marcellus Shale Region Illustrates Drivers for Pro-duced Water Desalination and Reuse. The Marcellus shaleplay is by far the largest shale gas reservoir in the United States,with estimated technically recoverable resources of approx-imately 140 trillion cubic feet of natural gas (140 Tcf or 4.0Tm3).3 Two “sweet spots” of drilling activity are northeasternand southwestern Pennsylvania, and the number of Marcellusshale well permits issued by the state has increased dramaticallysince 2007.24 The Marcellus play also underlies portions ofneighboring New York, where an effective moratorium on high-volume hydraulic fracturing operations was instituted in 2008and was recently extended to allow for additional health impactstudies.10

Management of high-salinity produced water from theMarcellus shale in Pennsylvania, which ranges in total dissolvedsolids concentration from 8000 to 360 000 mg/L with anaverage concentration of approximately 190 000 mg/L,25 hasevolved with changing regulations and the growth of the shalegas industry. During the period 2008−2011, the primarydisposal method has shifted from municipal wastewatertreatment facilities to industrial wastewater treatment facilitiesto underground injection wells and reuse for subsequent welldevelopment.14 Recent data from Pennsylvania show thatinternal reuse is now the most common produced watermanagement practice, and in 2012, 90% of produced water wasreused for hydraulic fracturing operations.13 The shift awayfrom treatment facilities that ultimately discharge to surfacewater in favor of other produced water management strategieshas been regulatory-driven in response to concerns aboutincreasing total dissolved solids (TDS) concentrations in thereceiving waters. In 2008, TDS concentrations in theMonongahela River rose to 900 mg/L, almost double theestablished water quality standard of 500 mg/L.26 The increasehas been partly attributed to shale gas produced water disposalat municipal wastewater treatment facilities, which are notequipped to remove dissolved solids. The high TDS

Figure 2. Interrelated shale play, economic, regulatory, and infrastructure considerations that influence decisions about produced water managementvia disposal or internal reuse without desalination versus external reuse, which requires produced water desalination either on-site at the well locationor at an off-site centralized facility.



concentration resulted in guidance from the PennsylvaniaDepartment of Environmental Protection advising municipalwastewater treatment facilities discharging to the river to limitproduced water volumes to 1% of the facility daily influentvolume. The department also encouraged concerned consum-ers for whom the Monongahela is a source of drinking water touse bottled water for cooking and drinking.27

Several high-profile incidents of groundwater and surfacewater contamination by natural gas and hydraulic fracturingfluids also occurred in Pennsylvania in 2009−2010. Theseincidents occurred during well construction or as a result ofpoorly constructed wells,24 and they heightened publicawareness of the potential negative impacts of shale gasproduction. As a result, the state implemented changes inregulatory requirements for shale gas well construction andwastewater storage, treatment, and disposal.24 The mostsignificant regulatory changes were the requirement thatproduced water be treated at a centralized treatment facilitybefore discharge to surface water or to a municipal treatmentfacility and the establishment of a monthly average TDSconcentration limit of 500 mg/L in the discharge.28

The treatment requirements and TDS limit for producedwater discharge to surface water in Pennsylvania increaseddemand for underground injection disposal and promptedgreater internal reuse of produced water.14 The Marcellus shaleregion generally does not have favorable geology for under-ground injection wells,24 and only eight injection wells exist inPennsylvania for underground disposal of produced water.29

For comparison, there are approximately 12 000 such wellsoperating in Texas.29 For Marcellus shale operations in

Pennsylvania, underground injection of produced waterprimarily occurs in Ohio. Concerns that underground injectionis inducing seismic events in Ohio30 may further constrain thisproduced water management option.In the Marcellus shale region, shale gas producers are

currently managing most produced water through internalreuse without desalination to remove dissolved solids.However, this reuse strategy is only a temporary solution. Asshale gas production in the Marcellus shale play matures,opportunities to reuse produced water in developing new wellswill decline while produced water pumping from establishedwells will continue.13,14 Whenever produced water volumesexceed demand for internal reuse, producers in the region willbe driven toward external reuse opportunities, which requiredesalination of produced water.The Marcellus shale region is currently the testing ground for

developing and implementing technologies for produced waterreuse, and thus, we review desalination technologies suitable forthe produced water quality and economic, regulatory, andinfrastructure considerations of this and similar shale gasregions. The reviewed technologies have potential applicationin shale plays worldwide where limited produced water disposaloptions will drive desalination of produced water wheneverproduced volumes exceed demand for internal reuse. Suitabletechnologies must be capable of desalinating high-salinity feedwaters, have a low propensity for fouling, and be modular andscalable for on-site treatment at shale gas well sites. Ideally,produced water desalination technologies will use inexpensive,locally available alternative energy sources to reduce the energycosts of produced water treatment.

Figure 3. Map illustrating the locations of active shale plays in the contiguous United States and the ranges of produced water total dissolved solids(TDS) concentrations from the U.S. Geological Survey’s Produced Waters Database25 supplemented with recent produced water quality data fromthe Marcellus shale play.33 Histograms illustrate the range of produced water TDS concentrations for the Williston Basin (Bakken and GammonShale Plays), Powder River Basin (Mowry Shale Play), Permian Basin (Avalon-Bone Spring and Barnett-Woodford Shale Plays), and AppalachianBasin (Devonian, Marcellus, and Utica Shale Plays).



■ UNIQUE CHALLENGES OF SHALE GAS PRODUCEDWATER TREATMENT

Produced water is challenging to treat because of its complexphysicochemical composition, which may vary over the lifetimeof the shale gas well. Treatment technologies should bedesigned for these potential changes in produced water qualityover time. Constituents of concern in produced water31 includedissolved and suspended organics, measured as total oil andgrease; suspended solids, such as formation solids, corrosionand scale products, and bacteria; production chemicals, whichmay contain proppants, friction reducers, biocides, andcorrosion inhibitors from the hydraulic fracturing fluid;12

naturally occurring radioactive material, specifically bariumand radium isotopes; and total dissolved solids (TDS),including hardness and heavy metals.Reducing the TDS concentration is the primary consid-

eration for treating produced water to a quality suitable fordischarge or for external reuse. Significant pretreatment toreduce fouling and scaling potential is required before producedwater may be desalinated using membrane or thermaltechnologies. In practice, multiple technologies that targetremoval of different constituents are often combined incommercial produced water treatment systems.31,32

Produced water TDS concentrations vary widely accordingto the geographic location and geologic basin from which theproduced water originates. Despite the relatively recentdevelopment of shale gas production, the hydrocarbon-producing basins in which shale plays are located are well-studied from conventional oil and gas operations. Geochemicalevaluation of produced water from Marcellus shale gas wellsreveals a common process of origin and similar characteristicsfor the shale gas produced water and produced water fromconventional oil and gas wells in the Marcellus shale region.33

Studies have concluded that the inorganic characteristics ofproduced water from natural gas production and from gasrecovery from underground storage are controlled primarily bythe geology of the surrounding rock formation.34,35 Con-sequently, the U.S. Geological Survey’s Produced WatersDatabase25 for oil and gas production wells in establishedhydrocarbon-producing basins may be used to describe thegeneral quality of produced water from major shale gas plays inthe United States. Figure 3 shows the locations of major shalegas plays in the contiguous United States and the ranges ofproduced water TDS concentrations for individual wells andselected hydrocarbon-producing basins from the database.Produced water TDS concentrations reported in the databaserange from 1000 to 400 000 mg/L.The TDS concentration of produced water constrains the

selection of appropriate desalination technologies. Reverseosmosis is typically employed to treat saline water, such asseawater, with TDS concentrations of up to approximately35 000 mg/L,36,37 which results in a feed solution TDSconcentration of approximately 70 000 mg/L for desalination at50% recovery. The hydraulic pressure required to overcome theosmotic pressure of high-salinity solutions can exceed theallowable pressure of the reverse osmosis membrane modulesand other process equipment.36,37 Consequently, high-salinitywaters with TDS concentrations greater than approximately35 000 mg/L, representing a TDS concentration of 70 000 mg/L at 50% recovery, must be desalinated by more energyintensive thermal technologies.36,37 Conventional thermaldesalination technologies, such as multistage flash and multipleeffect distillation, are well established. However, the highinvestment costs38 and the significant energy requirements andassociated energy costs of these technologies36 limit theirimplementation. The comparatively larger footprint and morecostly materials and equipment of conventional thermal

Figure 4. Schematic diagram of the mechanical vapor compression (MVC) process. After preheating, the feed stream (mpw) is directed to theevaporator, where it is mixed with the brine accumulating in the evaporator sump. A fraction of the brine stream is bled from the system (md) whilethe remainder (mr) is recirculated to the top of the evaporator unit to be sprayed over the outer surface of heat exchange tubes, over which waterevaporates. A vapor compressor suctions the saturated vapor (mp at Tpw,evap), discharging it as superheated steam into the heat transfer bundle (atTc,out > Tpw,evap). As the superheated steam flows inside the heat exchange tubes, it condenses (yielding distillate or product stream, mp at Tcond) andexchanges heat with brine trickling down the tubes’ outer surface (see inset). The principal energy input into the system is in the form of electricalenergy required to drive the vapor compressor (WVapor Compressor), and the feed, distillate, and recirculation pumps (WFeed, WDistillate, and WRecirculation,respectively).



desalination technologies compared to membrane desalinationtechnologies38 also imply that they are less mobile and scalable.

■ TECHNOLOGIES FOR PRODUCED WATERDESALINATION AND REUSE

Mechanical vapor compression, membrane distillation, andforward osmosis are three desalination technologies for high-salinity brines that are appropriate for the produced water inthe Marcellus shale region as well as other shale gas playsworldwide where conditions promote external reuse. Asdescribed herein, these technologies can achieve stringentpermeate TDS concentration limits for external reuseopportunities, their modular nature is conducive to theinfrastructure constraints of on-site treatment at shale gaswell sites, and their energy requirements and associated costsare competitive compared to more conventional thermaltechnologies. Mechanical vapor compression is a relativelywell-established technology, while membrane distillation andforward osmosis are emerging technologies that show promisefor low-energy desalination of high-salinity water. We present aprocess overview of the three technologies and discuss theadvantages and disadvantages of each for produced waterdesalination.Mechanical Vapor Compression. Mechanical vapor

compression (MVC) is a separation process that uses electricalenergy to supply thermal energy for desalination. MVC is beingused in seawater desalination as well as the desalination ofproduced water from heavy oil fields that employ steam-assistedgravity drainage for oil recovery.39−41 MVC consists of anopen-loop heat pump, where a compressor, driven by anelectrical motor, supplies the energy required to evaporatewater from a high-salinity feed.39 A flow diagram of the MVCprocess is shown in Figure 4.In an MVC system, the feed stream to be desalinated is

preheated by two plate heat exchangers using the heat providedby the distillate and brine streams leaving the system. The feedis then pumped to an evaporator-condenser, where it is mixedwith the brine accumulating at the sump of the evaporator.Some of this slurry is bled from the system to removeaccumulated dissolved solids. A recirculation pump conveys theremainder of the slurry to a nozzle distribution system, whichevenly spreads the slurry over a bundle of heat transfer tubes toform a continuous falling film. Evaporating water is suctionedby the vapor compressor, which discharges superheated steamto the heat transfer tubes. As it passes through the heat transfertubes in the evaporator-condenser, the superheated steamcondenses, providing the enthalpy of vaporization for the slurrytrickling down the outer surfaces of the tube bundle.40 Thefalling film pattern results in high heat transfer rates, therebyminimizing the compression ratio and the work to drive thecompressor.42 Both condensate and brine streams are pumpedout of the evaporator-condenser before being discharged orrecycled. Noncondensable gases are removed from theevaporator through a secondary condenser and exhaustedthrough a rotary vacuum pump.39,40 Typical operatingconditions for MVC seawater desalination units are anevaporator operating temperature of 60 °C and a compressionratio of 1.1.40

An alternate MVC system configuration, which has beenadopted to optimize the heat transfer coefficient, usesevaporators with a vertical tube bundle. In this configuration,the feed stream forms a falling film inside the heat exchangertubes where the water vaporizes. The heat of vaporization is

provided by superheated steam from the compressor, whichflows over the external surface of the heat transfer tubes andcondenses as distilled water.43

Both the heavy oil production and seawater desalinationindustries have adopted MVC technology, with more than 200MVC units commissioned worldwide.41 In heavy oil fieldsemploying steam assisted gravity drainage, MVC is regarded asthe standard method for produced water desalination andsubsequent reuse for steam generation using drum boilers.43

MVC has been used to treat produced water from German andDutch heavy oil fields.39 After separation of the extracted hotoil and produced water, MVC is used to desalinate theproduced water fraction, generating distillate and brine streams.The distillate is used as the feed for boiler water in the steamrecovery process, thus eliminating the need for reinjection ofthe produced water to the depleted oil reservoir.39 Numerousheavy oil fields have implemented MVC for the treatment andreuse of produced water, replacing traditional methodsincluding warm lime softening, filtration, and weak acid cationion exchange.43

The process performance of MVC units varies slightlydepending on the application. An MVC unit deployed in heavyoil fields to desalinate high-salinity produced water (TDSconcentration of 64 000 mg/L) for use as boiler feed waterexhibited a distillate capacity of 600 m3/day (4000 bbl/day) at30% recovery (mass of distillate per mass of feed).39 Thespecific energy consumption was 13.6 kW h/m3 distillate. MVCseawater desalination units treating seawater streams (TDSconcentration of approximately 38 000 mg/L) are capable ofproducing 500 m3/day (3000 bbl/day) of distilled water at anenergy consumption of 10.4−11.2 kW h/m3 distillate whenoperating at 40% recovery.40 Process performance (i.e., mass ofdistillate per unit energy input) increases with capacity,46 andgiven the weak dependence of vapor pressure on NaClconcentration,44 process performance appears to decreaseonly slightly with feed salinity45 at brine concentrations similarto seawater.46

Use of MVC for TDS removal presents a number ofadvantages compared to other technologies. In heavy oilrecovery, produced water desalination by MVC has reducedtreatment complexity and waste stream emissions compared totraditional treatment processes involving deoiling, softening,filtration, and ion exchange.43 In addition, capital costs arelower due to the modularized design of the evaporator, andoperating costs are reduced compared to traditional pro-cesses.43 The modular nature of the technology also impliesthat capacity can be increased by addition of evaporator-condenser units.40 Recycling desalinated water as boiler feed forheavy oil recovery also eliminates the need for disposing of theproduced water, which can be produced at four times thevolume of oil recovered.39

Compared to membrane-based seawater desalination tech-nologies, MVC is advantageous in that it is less prone to foulingby oil and grease, and therefore, less extensive pretreatmentmay be necessary prior to MVC desalination.40,43 Feedpretreatment by deoiling and antiscalant dosing (to preventcorrosion of and precipitation on the heat exchangers) is typicalof MVC units.40,43 The solubility of NaCl in water(approximately 370 000 mg/L at 60 °C)47 imposes an upperTDS concentration limit near 200 000 mg/L for MVC feedwater.48 Finally, when volatile organic compounds are presentin the product stream, secondary treatment (e.g., pervapora-tion)49 will be required.



The primary disadvantage of MVC compared to otherproduced water desalination technologies is its relatively highenergy requirement of 10.4−13.6 kW h/m3 distillate, dueprimarily to the electric power required to drive the vaporcompressor.39,40 Unlike other desalination technologies, such asmembrane distillation or forward osmosis, MVC requires high-grade electrical energy to perform the separation. This, in turn,necessitates access to an existing power grid or othercontinuous supply of electricity.50,51 Improvements in theheat transfer coefficient (e.g., through dropwise condensation)could decrease the energy requirements and associatedoperating costs.52

Membrane Distillation. Membrane distillation (MD) is anemerging technology that can utilize low-grade heat to driveseparation. In MD, the aqueous feed stream is separated fromthe permeate by a hydrophobic, microporous membrane.Liquid is unable to penetrate the membrane pores due to thehydrophobic nature of the membrane, and a difference in thepartial vapor pressure drives the transport of water vapor acrossthe membrane pores (Figure 5A).53−58 Previous studies havefocused on four main configurations: direct contact, air gap,sweeping gas, and vacuum MD. These configurations differprimarily in the way the vapor pressure gradient is maintainedand the condensation method for the vapor permeate.53−55,57,58

In addition to being the simplest configuration, direct contactMD has been studied most extensively and is identified as themost suitable configuration for purification of feed streams withnonvolatile solutes, such as desalination applications.57 Hence,for straightforwardness, we discuss only direct contact MD, butthe identical principles allow other MD configurations to besimilarly applied for produced water desalination.In direct contact MD, the temperature difference between a

warm feed solution and an ambient temperature aqueouspermeate stream drives water vapor flux.53−55,57,58 Ions,colloids, and macromolecules are nonvolatile, and hence, theyare retained in the feed, yielding pure water as the permeateproduct.55,57−59 Because the logarithm of the vapor pressure islinearly proportional to the reciprocal temperature (Clausius−Clapeyron relation), conducting MD at a higher feed solutiontemperature will result in a larger driving force and willgenerate exponentially greater vapor fluxes (Figure 5D, solidblue line), improving productivity.53,55,57,60 Additionally, theprocess efficiency is enhanced at higher feed temperatures, asillustrated by the dashed green line in Figure 5D, whichrepresents the fraction of thermal energy that is utilized forvaporizing water (i.e., permeate production) instead of beingdissipated by conductive heat loss across the membrane.60,61

Water flux rates in MD are only slightly sensitive to the feedsalinity.57,58,62 For example, a study found that increasing theTDS concentration of the feed from 35 000 to 75 000 mg/Lreduces the resultant permeate flux by only 5%.60 This featuremakes MD particularly suited to desalinate high-salinitysources, such as produced waters, without incurring substantialproductivity penalties. Almost complete salt rejection hasconsistently been reported in the literature, even for very highsalinity feeds.63−65 Compared to pressure-driven membraneprocesses, such as reverse osmosis, MD has a lower foulingpropensity due to larger membrane pores and the absence of anapplied hydraulic pressure.53,55,57,58 The relatively smallphysical footprint and modular nature of MD affords furtherversatility in adapting the technology to the unique infra-structure challenges of shale gas well sites.57

The composition of produced water can pose distinctivechallenges to MD. Small organic compounds and dissolvedgases that exert partial vapor pressures comparable to or higherthan water are transported across the membrane with the watervapor flux, causing contamination of the permeatestream.53,55,56 Certain feed components, such as alcohols andsurfactants, can lower the liquid surface tension of the feedsolution and cause wetting of the membrane pores (Figure5B).53,56,57 The feed solution can then flow directly across themembrane through the wetted pores, deteriorating permeatequality. To restore the vapor−liquid interface at the pores, thewetted membrane must be taken out of operation and driedcompletely, resulting in process downtime.53 Although foulingis reduced in MD compared to conventional pressure-drivenmembrane processes, it can, nonetheless, have detrimentalimpacts on performance. Fouling clogs membrane pores, whichleads to flux decline and pore wetting and imposes additionalhindrance to heat and mass transfer (Figure 5C). Mineralscaling is anticipated to be an important detrimental

Figure 5. Conceptual illustration of membrane distillation (MD). (A)The temperature difference between the warmer feed and coolerpermeate streams gives rise to a vapor pressure gradient which drivesthe transport of water vapor across the membrane pores. (B) Wettingof the pores, where the aqueous solution penetrates the hydrophobicmembrane, results in the feed solution flowing directly across themembrane, causing permeate quality to deteriorate. (C) Fouling of theMD membrane leads to performance decline, such as flux decay. (D)Permeate flux (blue solid line, left vertical axis) and thermal efficiency(green dashed line, right vertical axis) as a function of the feedtemperature for a typical commercial polypropylene membrane, withpermeate temperature set at 20 °C.60



phenomenon given the high salinity of the produced water feedsolution and the presence of salts close to their saturationconcentrations.55−57,66−68 The deleterious effects of particulate,organic, and biological fouling can also diminish processproductivity.53,55,56,62,67,69

Pretreatment to remove foulants and periodic membranecleaning will likely be necessary to maintain the productivity ofdesalination by MD.62,65,67,70−72 Pretreatment should also bedesigned to remove components from the produced water feedthat induce membrane pore wetting. Using membranes withsmaller pore sizes and more hydrophobic material can also helpminimize such facilitated wetting.53−58 To meet product waterquality standards, post-treatment might be required to removevolatile compounds and gases transported into the permeatefrom the feed solution.64 Pre- and post-treatment steps, as wellas membrane cleaning, will be important considerations in theimplementation of MD technology as they levy additionalchemical and energy costs on the overall process.57,60 Due tothe variability in the produced water composition, it isforeseeable that pre- and post-treatment protocols, cleaningstrategies, and membrane design will have to be tailored to thespecific water quality of the produced water source.57

The minimum energy required to desalinate produced waterin a hypothetical MD system without energy recovery devices isequivalent to the enthalpy of vaporization of the high-salinityfeed stream (approximately 680 kW h/m3). For the process tobe energetically viable, energy recovery devices such as heatexchangers will be necessary to reuse the thermal energy.However, MD is uniquely positioned to leverage the inherentlyelevated temperatures of the produced water feed solutions todrive the separation,22,73 and thus, it has the potential to be alow energy input desalination technology. Produced watertemperature could be as high as 100 °C, as discussedsubsequently in the Sources of Energy for Produced WaterDesalination section of this review. Pairing produced water thatis already at a temperature between 60 and 80 °C with apermeate stream at 20 °C can yield operationally practicalwater fluxes and thermal efficiencies (Figure 5D). Thermalenergy of the warm produced water is utilized to provide thedriving force for separation, and as such, the process requiresonly auxiliary inputs, such as electricity to circulate thesolutions, and chemical and energy costs for pre- and post-treatment steps.57,60,61 In cases where the produced water is notsufficiently hot to generate practical fluxes, additional thermalenergy can be supplied to elevate the feed solution temperatureand enhance productivity. This supplementary energy input canbe partially recovered by using heat exchangers to capture thethermal energy of the heated permeate stream leaving the MDmodule.54,74,75 A recent study estimates that operating MDwith heat recovery (with feed solutions initially at ambienttemperature) can potentially lower the thermal energyconsumption to approximately 40 kW h/m3 of product water(approximately 140 kJ/kg).60

The ability to utilize the inherent low-grade heat of the feedstream to produce operationally adequate water fluxes affordsMD unique opportunities to be an attractive, low-energytechnology for shale gas produced water desalination. However,MD will be challenged by the geospatial variability and thecomplexity of the produced water composition. The potentialapplication of MD to desalinate produced water has been thetopic of recent investigations,64,65 but more studies arenecessary to further our understanding of the potentialdifficulties, develop suitable membranes, optimize operating

parameters in membrane modules, and formulate effective pre-and post-treatment strategies in order to the advance theimplementation of MD.

Forward Osmosis. Forward osmosis (FO) removes TDS ina membrane-based separation process. Similar to reverseosmosis desalination, and unlike MVC or MD, FO achievesdissolved solids removal without the liquid permeate under-going a phase transition. Permeate water passes across a highlyselective membrane that retains dissolved solutes in the feedwater. The primary difference between forward and reverseosmosis is the driving force for separation. In reverse osmosis,hydraulic pressure is applied to overcome the feed waterosmotic pressure. However, in FO, water flux is driven by anosmotic pressure difference between the feed solution and aconcentrated draw solution on the permeate side that has ahigher osmotic pressure than the feed. The flow of permeatefrom the feed to the draw dilutes the draw solution and reducesthe osmotic pressure gradient that drives separation. Con-sequently, an additional draw solution regeneration step isrequired for separation of the draw solute from the permeateand reconcentration of the diluted draw solution.76

Forward osmosis has several advantages compared to othertechnologies for desalinating high-salinity feed waters. Becauseit is driven by osmotic pressure rather than hydraulic pressure,FO can desalinate high-salinity feed waters using simple andinexpensive low-pressure equipment. Unlike reverse osmosis,FO is not limited by the high-pressure operating limit thatcorresponds to a solution TDS concentration of approximately70 000 mg/L.36,68,77

Low-pressure operation of FO also translates to a relativelylow propensity for irreversible fouling of the membrane.78−80

Organic foulants that accumulate on the membrane surfacetend to be less compact than similar fouling layers in apressurized system, and the fouling layers can be more easilyremoved. This low fouling propensity can improve the overallFO process efficiency by reducing the pretreatment require-ments for produced water and their associated energy andcosts. In addition, low fouling propensity is expected to requireless aggressive and frequent membrane chemical cleanings,leading to longer membrane life and reduced replacement costs.Another specific advantage of FO for shale gas produced waterdesalination is the modularity of the treatment system, whichallows the capacity to be scaled up with the addition ofmembrane modules.The selection of an appropriate draw solute is crucial for FO

process efficiency. The draw solution must generate highosmotic pressure and be separated and reconcentratedefficiently with low energy costs. These restrictions are morepronounced when using FO to desalinate shale gas producedwater because the high-salinity produced water requires a drawsolution with a very high osmotic driving force, and the remoteand temporary nature of shale gas well sites may limit theavailable energy sources for draw solution regeneration. Drawsolutes that cannot generate a sufficiently high osmoticpressure, such as magnetic nanoparticles,81 stimuli-responsivepolymer hydrogels,82 and polyelectrolytes83 are not suitable forFO desalination of high-salinity produced water.Dissolved salts are the preferred draw solute as they can

generate high osmotic pressures and be rejected by themembrane. In a closed-loop system, the draw solution ofdissolved salts must be regenerated. Because the draw solutionmust have higher osmotic pressure than the feed, reverseosmosis cannot be used to reconcentrate the draw solution



when the feed salinity is higher than the correspondingtechnical pressure limitation of reverse osmosis. Consequently,previously suggested draw solutes that rely on reverse osmosisfor regeneration, such as MgSO4

84 and NaCl,68,85 are notapplicable for produced water treatment.A draw solution of dissolved salts may also be treated as a

sacrificial draw solution,86 meaning that the diluted drawsolution will be used as-is rather than being regenerated andreturned to the membrane. In this configuration, water istransferred from the feed to the draw so that FO will essentiallyserve as a pretreatment for the feed stream. The sacrificial drawapproach is the basis for the Green Machine by HydrationTechnology Innovations.87 The Green Machine uses aconcentrated NaCl draw solution to treat produced water.Following treatment, the diluted NaCl draw solution, whichremains at a TDS concentration greater than or equal to theproduced water feed, is recycled for well developmentactivities.87

Thermolytic salts, which can change phase via a change insolution temperature, are the most suitable FO draw solutecandidates for desalinating high-salinity produced water.Thermal energy can be used in draw solution regeneration tovolatilize the thermolytic solutes, recover the permeate water,and reconcentrate the draw solution. A thermal draw solutionrecovery process does not have the restriction of a maximumoperating hydraulic pressure. Therefore, it can achieveseparation of the draw solute from the permeate water evenin highly concentrated draw solutions. When an inexpensive,low-grade heat source is available to fuel the thermal separation,it may be possible to significantly reduce the thermal energycosts,88 such as using a geothermal source for produced waterdesalination.Currently, ammonia−carbon dioxide is the most mature

thermolytic draw solute in terms of research and develop-ment.77,89−91 An ammonia−carbon dioxide draw solutioncontains ammonium salts that can generate osmotic pressuresgreater than 200 atm89 and be decomposed to ammonia andcarbon dioxide gases at a relatively low temperature(approximately 60 °C at atmospheric pressure).77 In an FOsystem, the ammonia and carbon dioxide gases are capturedand condensed to reconcentrate the draw solution. Seawaterdesalination by ammonia−carbon dioxide FO has been widelydiscussed because of the high osmotic pressure of the draw

solution, the facile draw regeneration procedure, and thepotential to use low-cost thermal energy for draw soluterecovery.90,91 The ability of the ammonia−carbon dioxide drawsolution to generate significant water fluxes when desalinating ahigh-salinity feed solution (TDS concentration of 140 000 mg/L) was demonstrated in early research.89

Recently, a pilot-scale operation of an FO system usingammonia−carbon dioxide draw solution demonstrated thepotential for desalinating shale gas produced water (Figure6A).92 The feed was a shale gas produced water from theMarcellus shale region with an average TDS concentration of73 000 mg/L. Due to the high-salinity feed and the targetedhigh water recovery, a reverse osmosis step was used as asecond desalination pass for the permeate after FO in order toachieve a permeate TDS concentration lower than 500 mg/L.However, this secondary RO step may not be necessary whenconsidering less strict salinity standards for produced waterdischarge. A distillation column and condenser were used toregenerate the draw solution. The pilot successfully achieved64% water recovery, desalinating produced water to achieve apermeate TDS concentration less than 300 mg/L. In addition,the permeate had a very low concentration of total organiccarbon (TOC), and the concentrations of major cations andanions reached potable water values (Figure 6B). The resultingconcentrated feed water had an average TDS concentration of180 000 mg/L. Although reported water fluxes from the pilotstudy were relatively low (approximately 3 L/(m2 h)), furtheroptimization and development in both membrane de-sign76,86,93,94 and in module hydrodynamics95 are expected toimprove the flux.An energy analysis of the pilot ammonia−carbon dioxide FO

system showed that 97% of the consumed energy was thermal,and the remainder was electrical. The specific energyconsumption of the FO pilot system was shown to besignificantly lower (2.3 times less) than other thermaldistillation methods. This is due primarily to the fact that, intraditional thermal distillation methods, the feed water isevaporated at a latent heat of evaporation of 2260 kJ/kg, whilein forward osmosis, smaller volumes of ammonia and carbondioxide are evaporated at latent heats of vaporization of 1369and 574 kJ/kg, respectively.96 A suggested opportunity foradditional significant energy savings is recovering the energyreleased during condensation of the ammonia and carbon

Figure 6. (A) Schematic illustration of a forward osmosis (FO) pilot system for produced water desalination (adapted from ref 92). (B) Productwater concentrations (bars, left vertical axis) and corresponding percent rejection (symbols, right vertical axis) of total dissolved solids (TDS), totalorganic carbon (TOC), and representative major cations and anions in the for an FO pilot-scale system desalinating produced water (data from ref92). The “BDL” next to TOC means “below detection limit”.



dioxide gases, in a similar manner to the recovery of the watervapor latent heat in MVC.92 In an analysis including energyrecovery, the ammonia−carbon dioxide FO system had aspecific energy consumption 42% less than a single-stage MVCsystem operating under the same conditions (21 and 37 kW h/m3 of permeate, respectively).92 Another potential opportunityfor energy savings is using the heat of elevated-temperatureproduced water to help recover ammonia and carbon dioxidegases to regenerate the draw solution.Despite the potential of the ammonia−carbon dioxide draw

solution, it suffers from ammonia loss due to reverse solute fluxthrough the FO membrane from the draw to the feed side. Oneway to resolve the ammonia reverse flux is stripping theammonia from the feed stream after FO treatment in order torecycle it and to minimize the draw solute loss (Figure 6A). Inthe discussed pilot-scale system, the reported ammoniaconcentration in the product water after stripping was lessthan 1 mg/L.92 Other ways to improve the reverse draw soluteflux involve the development of membranes that are moreselective against ammonia or identifying other effectivethermolytic draw solutes.Trimethylamine (TMA) may be a promising thermolytic

draw solute alternative to ammonia.97 Its thermodynamicproperties suggest comparable volatility to ammonia, and thereverse draw solute flux of TMA would be reduced comparedto ammonia because the larger TMA molecule is better rejectedby the FO membrane. Switchable polarity solvents (SPSs)98,99

may be another alternative to ammonia−carbon dioxide. SPSsare a class of relatively large amine solvents that change theirpolarity and, thus, their miscibility in water, upon pH changeinitiated by the introduction or removal of carbon dioxide insolution.While TMA and SPSs show potential as draw solutes for FO,

they are still at early stages of investigation. However, a pilot-scale FO system using an ammonia−carbon dioxide drawsolution has demonstrated the ability to treat high-salinity shalegas produced water using less energy than conventional thermalmethods. The relatively low specific energy consumption ofFO, combined with advantages such as low propensity forirreversible fouling, modularity, and the potential to use low-grade heat as the energy source, indicate that FO can be apotential desalination technology for external reuse of high-salinity shale gas produced water.

■ SOURCES OF ENERGY FOR PRODUCED WATERDESALINATION

The reviewed technologies for desalination of produced waterinvolve significant consumption of energy. As thermally drivenseparation processes, their efficiency is inherently low.100 Thislow energy efficiency is unavoidable in produced waterdesalination because nonthermal processes with higherefficiencies cannot desalinate such high-salinity feed waters.The potential location of these treatment processes at well sites,often away from substantial infrastructure, suggests thatsupplying energy may be a critical challenge to the widespreaddesalination of produced water for external reuse. We thereforeexamine considerations and options for feasibly powering thesetechnologies with available resources according to their thermaland electrical energy needs. As some debate currentlysurrounds the claim that natural gas represents a lower-carbonenergy source,101,102 we give special attention to the environ-mental implications of using each potential energy resource.

The modular construction, scalability, and lack of emissionsassociated with solar energy make it immediately appealing forgeneration of electrical or thermal energy, either directly orfrom solar electrical energy. The ability to transport solarenergy capturing equipment between well sites as producedwater flow rates fluctuate is another distinct advantage.However, solar power cannot be the primary energy sourcefor produced water desalination for several reasons. First, theintermittency of solar exposure would necessitate the use of analternative energy source to maintain a constant power supplyfor continuous operation. The footprint of a supply system,particularly for thermal energy, could also be prohibitively largewhen some concerns about disruption of current land usealready exist.103,104 In addition, the Marcellus shale regionexperiences relatively low levels of solar radiation, and theresulting payback period without incentives is decades long forboth photovoltaics and solar water heating systems.105

However, this payback period would shrink considerablywhen solar energy capture is compared with the defaultalternative for off-grid power production: diesel-poweredgenerators. Solar energy becomes competitive in the Marcellusregion against effective costs of $0.2−$0.3 per kW h or $3.5−$4per therm.105 At present, solar energy appears to be useful onlyas a supplemental source of electricity at Marcellus well sites. Itwould be better suited for other shale plays that receive moresolar radiation or under scenarios with enhanced economicincentive.Geothermal energy, however, would be readily and

continually available at well sites in the Marcellus shale region.The elevated temperatures of extracted gas and produced waterin this play and many others are a source of significant energy(Figure 7A). In the western United States, temperatures mayexceed 100 °C,106 facilitating electricity generation at the wellsite.107 The lower temperatures that prevail in the eastern U.S.are more suitable for direct thermal uses of energy. This energycould be easily exploited to power MD or FO desalination ofproduced water, at least in part. While this energy sourcecannot be shifted between well sites to address spatiotemporalvariability in demand for produced water desalination energy, assolar energy production could be, the relative simplicity of itscapture gives it a significant advantage. If additional heat isrequired, a geothermal well could be constructed at the well siteusing essentially the same expertise and equipment used to drillthe gas-producing wells. Drilling deeper will also yield highertemperatures, though at increased initial cost.Although temperature data at typical hydraulic fracturing

depths are not currently available across the Marcellus shale,estimates can be made. Figure 7 maps temperatures 3.5 km (2.2mi.) below the ground surface (bgs), based on measured fielddata and some modeling.106 Marcellus shale gas wells will notreach this depth or temperature; their maximum depth will beapproximately 2.7 km (1.7 mi).24 The general rule based onshallower water well development, an increase of 1 °F (0.6 °C)per 100 ft of depth,108 likely underestimates the temperature ofproduced water within the Marcellus shale. In southwestern ornortheastern Pennsylvania, this guideline would estimate 57 °Ctemperatures for the deepest shale gas wells. However, the rulewould give 72 °C at 3.5 km bgs, a significant underestimation ofthe more data-based temperature range of 100−125 °Cindicated for those regions in Figure 7. Therefore, the truemaximum temperature in the Marcellus, at its deepest wells, islikely 60−100 °C. Thermal energy at this temperature would behighly useful for the desalination technologies reviewed herein.



Further development of geothermal resources could also takeplace in a location central to several wells to balance the cost oftransport to a central, but still nearby, treatment facility with thebenefit of aggregating produced water flow rates and quality todampen fluctuations in the feed stream. By siting geothermal-energy wells closer to existing commercial or residentialbuildings, a long-term benefit of this development could berealized for the local population. After the cessation of naturalgas exploration and production in the area, the geothermalenergy resource could remain available to the community forsustainable and low-cost temperature control for buildings.Thus, the oil and gas industry would be helping to provideclean sources of energy for the future while also working tosupply adequate energy for today’s population in the context oftoday’s infrastructure.While clean, renewable energies can significantly contribute

to the energy requirements of produced water treatment, somesituations will inevitably require more conventional on-demand

energy. Under certain economic and regulatory drivers, it maybe advantageous to utilize a portion of the just-producednatural gas at the well site as an energy supply rather than dieselfuel. This would reduce direct emissions, eliminate thetransportation of fuel to the well site, and conserve the costsof processing the natural gas for long-distance distribution,including the potential loss of natural gas through leaks inpipelines. Processing of natural gas typically includes someprocessing at the well site and additional processing at acentralized facility. Water, acid, unsuitable gases, andcondensates are removed to enhance the product heatingvalue and to protect the distribution system. In shale playswhere the gas has little acid, as anecdotal evidence suggests maybe the case in the Marcellus, it may be feasible to burn just-produced, minimally processed natural gas to boost thermalenergy on an as-needed basis. This possibility highlights againthe importance of regulatory and economic drivers indetermining the fate of produced water; the gas industry willnot consume their primary product in desalinating water forreuse without significant incentive.The current state of technology and research does not

support exhaustive and direct comparisons between thesetechnologies in terms of their energy use profile. Nevertheless,some comparisons can be made and are included in Table 1. Asthese technologies mature, particularly for FO and MD, energyconsumption will be reduced and better understood. Theevolution of nontechnical drivers will heavily influence therelative attractiveness of both these energy resources andproduced water desalination technologies.

■ OUTLOOKShale gas production is expanding in the United States andworldwide, despite controversy surrounding the environmentaland public health implications of shale gas development andmanagement of the resulting produced water. In the Marcellusshale region, and other shale plays worldwide with similarconstraints, the anticipated combination of contracting internalreuse opportunities and economic, regulatory, and infra-structure drivers for external reuse will move the shale gasindustry to desalinate produced water. Technological innova-tion is needed now to advance existing methods and developnew, energy-efficient technologies suitable for desalinating high-salinity produced water. Produced water desalination forexternal reuse is technologically feasible. Advancing this reusein a safe and economical manner is the challenge for shale gasproducers, engineers, researchers, and regulators.We have reviewed MVC, MD, and FO as three suitable

technologies for desalinating high-salinity produced water.Though they are technically suitable, the capital and operatingcosts of these technologies must be reduced to facilitate their

Figure 7. Temperature 3.5 km (2.2 miles) below the ground surface(data from ref. 106) in (A) the contiguous U.S., with major shale playsoutlined (data from ref. 109), and (B) the northeastern U.S., with theMarcellus shale play outlined. The Marcellus shale is found at depthsof up to about 2.7 km.24 Data displayed in Google Earth (Data SIO,NOAA, U.S. Navy, NGA, GEBCO; Image USGS Copyright 2013CnesSpot Image; Image Copyright 2013 TerraMetrics; Copyright2013 Google).

Table 1. Expected Use and Source of Electrical and Thermal Energies to Power the Desalination Technologies Revieweda

energy use potential energy sources

technology electrical thermal (temperature) most sustainable conventional (to supplement)

MVC high noneconverted from electricity solar photovoltaic, electricity from high temperaturegeothermal

connection to grid or diesel-generatedelectricity

MD low high (60−80 °C, higher flux if highertemperature)

geothermal just-produced gas or diesel-generatedelectricity

FO low high (≥60 °C for NH3−CO2 drawsolution)

geothermal just-produced gas or diesel-generatedelectricity

aGeothermal energy can be harvested from wells drilled for that specific and exclusive purpose or from the elevated temperature of the producedwater.



economical implementation. Capital costs may be reduced withmodular and scalable systems that can be relocated or expandedto match the dynamic demand for produced water desalination.To this end, advancements in module design and operatingparameters, especially for less-mature MD and FO technolo-gies, may reduce costs. Reducing the extent of pretreatmentrequired for these desalination technologies and increasing theefficiency of pretreatment technologies themselves could alsoreduce the capital costs of produced water desalination.Significant savings could be realized by improving the hydraulicfracturing process to decrease the volume or reduce the TDSconcentration of produced water that must be desalinated.Operating costs of a desalination technology can be reduced

by lowering the specific energy consumption of the separationprocess (the amount of energy required to desalinate a unitvolume of feed water) or by changing the type of energyemployed in the separation. Using low-quality but inexpensiveenergy, such as geothermal energy, is one effective means ofreducing the operating costs. Of the technologies we havereviewed, the specific energy consumption of MVC is the mostwell-established. MVC relies on relatively expensive, primeelectrical energy to drive the thermal separation process, butimprovements in the heat transfer coefficient of the evaporator-condenser could reduce these energy requirements.MD and FO are less established than MVC, but these

technologies have the potential to harness the elevatedtemperature of produced water to help drive the separation.Using abundant, inexpensive geothermal energy can signifi-cantly reduce the energy costs of MD and FO, thus increasingthe economic incentive for their use. Materials and processimprovements in MD and FO can further reduce the energyand operating costs of these technologies. Materials improve-ments include membranes that better reject draw solutes in FOand high-permeability, fouling-resistant membranes for bothFO and MD. Reduced propensity for fouling translates toreduced operating costs through less extensive system pretreat-ment and cleaning requirements. Process improvementsinclude integrating energy recovery devices to recover thermalenergy from the elevated-temperature permeate stream in MDand from the volatilized draw solutes in FO to improve theoverall energy efficiency. A recent pilot-scale demonstration ofproduced water desalination by FO is an important milestonetoward full-scale implementation of the technology. AlthoughMD has yet to be demonstrated at pilot- or full-scale, ourunderstanding of the technology has progressed considerablyfrom a recent flurry of research activity. A shift in the shale gasindustry toward produced water desalination for external reuseindustry may provide the incentives that lead to commercializa-tion of MD technology.Research in other areas can improve understanding of the

economics of produced water management and help incentivizedesalination for external reuse. More detailed assessments ofthe specific energy consumption of MVC, MD, and FO fordesalinating high-salinity produced waters are needed to makedirect comparisons between the technologies. Study ofproduced water management considerations and resultingpractices in different shale plays, such as the Barnett shaleand the Marcellus shale, can elucidate the specific drivers thatinitiate external reuse, and this information can informregulators and policy makers. Considering the effect of reusetechnologies on the overall water and energy balance of shalegas production will improve understanding of the potentialsavings and importance of this effort. More centralized and

accessible sources of data for produced water quality,temperature, volumes, and management practices will facilitatestudy of these topics.The horizontal drilling and hydraulic fracturing technologies

that make shale gas production a feasible and market-changingactivity are remarkable engineering achievements. The shale gasindustry overcame many difficult technical challenges toaccomplish the profitable recovery of gas where it waspreviously deemed inaccessible. Similarly, the technicalchallenges in desalinating shale gas produced water to enableits external reuse are surmountable, and MVC, MD, and FOrepresent promising technologies. In each shale play,economics, influenced by infrastructure and regulation, willdetermine how far these technologies must advance toaccomplish widespread reuse.

■ AUTHOR INFORMATION

Corresponding Author*E-mail: [email protected].

NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTS

This publication was developed under STAR FellowshipAssistance Agreement Nos. FP-91750001-0 and FP-91733801-0 awarded by the U.S. Environmental ProtectionAgency (EPA) to D.L.S. and L.H.A.C., respectively. It has notbeen formally reviewed by EPA. The views expressed in thispublication are solely those of the listed authors, and EPA doesnot endorse any products or commercial services mentioned inthis publication.

■ REFERENCES(1) United States Energy Information Administration. World ShaleGas Resources: An Initial Assessment of 14 Regions Outside the UnitedStates; United States Department of Energy: Washington, DC, April2011; pp 1−365.(2) International Energy Agency.World Energy Outlook 2011: Are WeEntering a Golden Age of Gas?; International Energy Agency: Paris,France, 2011; pp 1−131.(3) United States Energy Information Administration. Annual EnergyOutlook; United States Department of Energy: Washington, DC, 2012.(4) Helman, C. The Arithmetic of Shale Gas. Forbes June 22, 2012.(5) MIT Energy Initiative. The Future of Natural Gas; Cambridge,MA, 2011.(6) Osborn, S. G.; Vengosh, A.; Warner, N. R.; Jackson, R. B.Methane contamination of drinking water accompanying gas-welldrilling and hydraulic fracturing. Proc. Natl. Acad. Sci. U.S.A. 2011, 108(20), 8172−8176.(7) Jackson, R. B.; Vengosh, A.; Darrah, T. H.; Warner, N. R.; Down,A.; Poreda, R. J.; Osborn, S. G.; Zhao, K.; Karr, J. D. Increased straygas abundance in a subset of drinking water wells near Marcellus shalegas extraction. Proc. Natl. Acad. Sci. U.S.A. 2013, 110 (28), 11250−11255.(8) Olmstead, S. M.; Muehlenbachs, L. A.; Shih, J.-S.; Chu, Z.;Krupnick, A. J. Shale gas development impacts on surface water qualityin Pennsylvania. Proc. Natl. Acad. Sci. U.S.A. 2013, 110 (13), 4962−4967.(9) Howarth, R.; Santoro, R.; Ingraffea, A. Methane and thegreenhouse-gas footprint of natural gas from shale formations. Clim.Change 2011, 106 (4), 679−690.(10) Associated Press. NY fracking rules in limbo as deadline arrives;environmentalists urge opening health studies. The Washington Post,Feb 27, 2013.



mailto:[email protected]

(11) Havard, K. Fracking in Maryland: Still under study. TheWashington Post, Jan 17, 2013.(12) Gregory, K. B.; Vidic, R. D.; Dzombak, D. A. WaterManagement Challenges Associated with the Production of ShaleGas by Hydraulic Fracturing. Elements 2011, 7 (3), 181−186.(13) Vidic, R. D.; Brantley, S. L.; Vandenbossche, J. M.; Yoxtheimer,D.; Abad, J. D. Impact of Shale Gas Development on Regional WaterQuality Science 2013, 340 (6134).(14) Lutz, B. D.; Lewis, A. N.; Doyle, M. W. Generation, transport,and disposal of wastewater associated with Marcellus Shale gasdevelopment. Water Resour. Res. 2013, 49 (2), 647−656.(15) United States Energy Information Administration. Review ofEmerging Resources: U.S. Shale Gas and Shale Oil Plays; United StatesDepartment of Energy: Washington, DC, 2011.(16) Clark, C. E.; Veil, J. A. Produced Water Volumes and ManagementPractices in the United States. United States Department of Energy,Argonne National Laboratory: Argonne, IL, 2009; p 59.(17) Ground Water Protection Council, ALL Consulting. ModernShale Gas Development in the United States: A Primer; United StatesDepartment of Energy, Office of Fossil Energy: Washington, D.C.,2009; p 96.(18) Nicot, J.-P.; Scanlon, B. R. Water Use for Shale-Gas Productionin Texas, U.S. Environ. Sci. Technol. 2012, 46 (6), 3580−3586.(19) Murray, K. E. State-Scale Perspective on Water Use andProduction Associated with Oil and Gas Operations, Oklahoma, U.S.Environ. Sci. Technol. 2013, 47 (9), 4918−4925.(20) Mantell, M. E. Produced Water Reuse and Recycling Challengesand Opportunities Across Major Shale Plays. USEPA TechnicalWorkshops for the Hydraulic Fracturing Study: Water ResourcesManagement, Arlington, VA, May; United States EnvironmentalProtection Agency Office of Research and Development, 2011.(21) United States Government Accountability Office. Information onthe Quantity, Quality, and Management of Water Produced during Oiland Gas Production; Washington, D.C., 2012.(22) Interstate Oil and Gas Compact Commission, ALL Consulting.A Guide to Practical Management of Produced Water from Onshore Oiland Gas Operations in the United States; U.S. Department of Energy,National Petroleum Technology Office: Washington, D.C., 2006.(23) Shariq, L. Uncertainties Associated with the Reuse of TreatedHydraulic Fracturing Wastewater for Crop Irrigation. Environ. Sci.Technol. 2013, 47 (6), 2435−2436.(24) Marcellus Shale Advisory Commission. Report of the Governor’sMarcellus Shale Advisory Commission; Commonwealth of Pennsylvania:Harrisburg, PA, July 22, 2011; p 136.(25) Breit, G. N.; Otton, J. K. Produced Waters Database; UnitedStates Geological Survey: Reston, VA, May 29, 2002.(26) Renner, R. Pennsylvania to regulate salt discharges. Environ. Sci.Technol. 2009, 43 (16), 6120−6120.(27) Sapien, J. With Natural Gas Drilling Boom, Pennsylvania FacesFlood of Wastewater. Sci. Am. October 5, 2009.(28) Wastewater Treatment Requirements. Pennsylvania Code,Section 10, Title 25, Chapter 95, 2010.(29) McCurdy, R. Underground Injection Wells for Produced WaterDisposal. USEPA Technical Workshops for the Hydraulic FracturingStudy: Water Resources Management, Arlington, VA, May; United StatesEnvironmental Protection Agency, Office of Research and Develop-ment, 2011.(30) Ohio Department of Natural Resources. Preliminary Report onthe Northstar 1 Class II Injection Well and the Seismic Events in theYoungstown, Ohio, Area; Ohio Department of Natural Resources:Columbus, OH, 2012.(31) Fakhru’l-Razi, A.; Pendashteh, A.; Abdullah, L. C.; Biak, D. R.A.; Madaeni, S. S.; Abidin, Z. Z. Review of technologies for oil and gasproduced water treatment. J. Hazard. Mater. 2009, 170 (2−3), 530−551.(32) Drewes, J. E. An Integrated Framework for Treatment andManagement of Produced Water: Technical Assessment of Produced WaterTreatment Technologies: The Colorado School of Mines: Golden, CO,2009; p 158.

(33) Haluszczak, L. O.; Rose, A. W.; Kump, L. R. Geochemicalevaluation of flowback brine from Marcellus gas wells in Pennsylvania,USA. Appl. Geochem. 2013, 28, 55−61.(34) Johnson, B.; Kanagy, L.; Rodgers, J.; Castle, J. Chemical,Physical, and Risk Characterization of Natural Gas Storage ProducedWaters. Water, Air, Soil Pollut. 2008, 191 (1), 33−54.(35) Benko, K. L.; Drewes, J. E. Produced Water in the WesternUnited States: Geographical Distribution, Occurrence, and Composi-tion. Environ. Eng. Sci. 2008, 25 (2), 7.(36) Greenlee, L. F.; Lawler, D. F.; Freeman, B. D.; Marrot, B.;Moulin, P. Reverse osmosis desalination: Water sources, technology,and today’s challenges. Water Res. 2009, 43 (9), 2317−2348.(37) Fritzmann, C.; Lowenberg, J.; Wintgens, T.; Melin, T. State-of-the-art of reverse osmosis desalination. Desalination 2007, 216 (1−3),1−76.(38) Ghaffour, N.; Missimer, T. M.; Amy, G. L. Technical review andevaluation of the economics of water desalination: Current and futurechallenges for better water supply sustainability. Desalination 2013,309, 197−207.(39) Koren, A.; Nadav, N. Mechanical vapour compression to treatoil field produced water. Desalination 1994, 98 (1−3), 41−48.(40) Veza, J. M. Mechanical Vapor Compression Desalination Plants- a Case-Study. Desalination 1995, 101 (1), 1−10.(41) Zimerman, Z. Development of Large-Capacity High-EfficiencyMechanical Vapor Compression (Mvc) Units. Desalination 1994, 96(1−3), 51−58.(42) Mistry, K. H.; McGovern, R. K.; Thiel, G. P.; Summers, E. K.;Zubair, S. M.; Lienhard, J. H. Entropy Generation Analysis ofDesalination Technologies. Entropy 2011, 13 (10), 1829−1864.(43) Heinz, W. F. Is a Paradigm Shift in Produced Water TreatmentTechnology Occurring at SAGD Facilities? In Canadian InternationalPetroleum Conference (CIPC), Calgary, June 12−14, 2007.(44) Xu, T.; Pruess, K. Thermophysical properties of sodium nitrate andsodium chloride solutions and their effects on fluid flow in unsaturatedmedia; Lawrence Berkeley National Laboratory: Berkeley, CA, 2001.(45) Darwish, M. A. Thermal analysis of vapor compressiondesalination system. Desalination 1988, 69 (3), 275−295.(46) Bahar, R.; Hawlader, M. N. A.; Woei, L. S. Performanceevaluation of a mechanical vapor compression desalination system.Desalination 2004, 166 (1−3), 123−127.(47) Langer, H.; Offermann, H. On the solubility of sodium chloridein water. J. Cryst. Growth 1982, 60 (2), 389−392.(48) Khazhsetovich, K. R.; Fedorovich, M. B.; Antonovich, R. B.;Ivanovna, M. L.; Gennadievich, A. E.; Vladimirovna, F. O.;Grigorievich, N. E.; Pavlovich, V. V.; Ilich, G. S.; Davydovich, C. V.;Ekhilievich, S. M.; Sergeevich, D. A.; Nikolaevich, D. A.; Saferbievich,B. K.; Anatolievich, R. J. Method for complex processing of sea-water.United States Patent 5,814,224, 1998.(49) Lipnizki, F.; Hausmanns, S.; Ten, P.-K.; Field, R. W.;Laufenberg, G. Organophilic pervaporation: prospects and perform-ance. Chem. Eng. J. 1999, 73 (2), 113−129.(50) Lokiec, F.; Ophir, A. The Mechanical Vapor Compression: 38Years of Experience. IDA World Congress, Maspalomas, Gran Canaria,Spain, October 21−26, 2007.(51) Al-Juwayhel, F.; El-Dessouky, H.; Ettouney, H. Analysis ofsingle-effect evaporator desalination systems combined with vaporcompression heat pumps. Desalination 1997, 114 (3), 253−275.(52) Lukic, N.; Diezel, L. L.; Froba, A. P.; Leipertz, A. Economicalaspects of the improvement of a mechanical vapour compressiondesalination plant by dropwise condensation. Desalination 2010, 264(1−2), 173−178.(53) Lawson, K. W.; Lloyd, D. R. Membrane distillation. J. Membr.Sci. 1997, 124 (1), 1−25.(54) Alklaibi, A. M.; Lior, N. Membrane-distillation desalination:status and potential. Desalination 2005, 171 (2), 111−131.(55) Curcio, E.; Drioli, E. Membrane distillation and relatedoperations - A review. Sep. Purif. Rev. 2005, 34 (1), 35−86.



(56) El-Bourawi, M. S.; Ding, Z.; Ma, R.; Khayet, M. A framework forbetter understanding membrane distillation separation process. J.Membr. Sci. 2006, 285 (1−2), 4−29.(57) Khayet, M.; Matsuura, T. Membrane distillation: principles andapplications; Elsevier: Oxford, 2011.(58) Alkhudhiri, A.; Darwish, N.; Hilal, N. Membrane distillation: Acomprehensive review. Desalination 2012, 287, 2−18.(59) Lawson, K. W.; Lloyd, D. R. Membrane distillation. II. Directcontact MD. J. Membr. Sci. 1996, 120 (1), 123−133.(60) Al-Obaidani, S.; Curcio, E.; Macedonio, F.; Di Profio, G.; Ai-Hinai, H.; Drioli, E. Potential of membrane distillation in seawaterdesalination: Thermal efficiency, sensitivity study and cost estimation.J. Membr. Sci. 2008, 323 (1), 85−98.(61) Criscuoli, A.; Carnevale, M. C.; Drioli, E. Evaluation of energyrequirements in membrane distillation. Chem. Eng. Process. 2008, 47(7), 1098−1105.(62) Meindersma, G. W.; Guijt, C. M.; de Haan, A. B. Desalinationand water recycling by air gap membrane distillation. Desalination2006, 187 (1−3), 291−301.(63) Cath, T. Y.; Adams, V. D.; Childress, A. E. Experimental studyof desalination using direct contact membrane distillation: a newapproach to flux enhancement. J. Membr. Sci. 2004, 228 (1), 5−16.(64) Singh, D.; Sirkar, K. K. Desalination of brine and producedwater by direct contact membrane distillation at high temperatures andpressures. J. Membr. Sci. 2012, 389, 380−388.(65) Alkhudhiri, A.; Darwish, N.; Hilal, N. Produced water treatment:Application of Air Gap Membrane Distillation. Desalination 2013, 309,46−51.(66) Gryta, M. Long-term performance of membrane distillationprocess. J. Membr. Sci. 2005, 265 (1−2), 153−159.(67) Gryta, M. Fouling in direct contact membrane distillationprocess. J. Membr. Sci. 2008, 325 (1), 383−394.(68) Martinetti, C. R.; Childress, A. E.; Cath, T. Y. High recovery ofconcentrated RO brines using forward osmosis and membranedistillation. J. Membr. Sci. 2009, 331 (1−2), 31−39.(69) Gryta, M. The assessment of microorganism growth in themembrane distillation system. Desalination 2002, 142 (1), 79−88.(70) Hsu, S. T.; Cheng, K. T.; Chiou, J. S. Seawater desalination bydirect contact membrane distillation. Desalination 2002, 143 (3), 279−287.(71) Gryta, M.; Tomaszewska, M.; Karakulski, K. Wastewatertreatment by membrane distillation. Desalination 2006, 198 (1−3),67−73.(72) Gryta, M. Chemical pretreatment of feed water for membranedistillation. Chem. Pap. 2008, 62 (1), 100−105.(73) Ray, J. P.; Engelhardt, F. R. Produced water: technological/environmental issues and solutions; Plenum Press: New York, 1992; p616.(74) Fane, A. G.; Schofield, R. W.; Fell, C. J. D. The efficient use ofenergy in membrane distillation. Desalination 1987, 64, 231−243.(75) Kurokawa, H.; Sawa, T. Heat recovery characteristics ofmembrane distillation. Heat Transfer−Japn. Res. 1996, 25 (3), 135−150.(76) Cath, T. Y.; Childress, A. E.; Elimelech, M. Forward osmosis:Principles, applications, and recent developments. J. Membr. Sci. 2006,281 (1−2), 70−87.(77) McCutcheon, J. R.; McGinnis, R. L.; Elimelech, M. A novelammonia-carbon dioxide forward (direct) osmosis desalinationprocess. Desalination 2005, 174 (1), 1−11.(78) Lee, S.; Boo, C.; Elimelech, M.; Hong, S. Comparison of foulingbehavior in forward osmosis (FO) and reverse osmosis (RO). J.Membr. Sci. 2010, 365 (1−2), 34−39.(79) Mi, B. X.; Elimelech, M. Organic fouling of forward osmosismembranes: Fouling reversibility and cleaning without chemicalreagents. J. Membr. Sci. 2010, 348 (1−2), 337−345.(80) Tiraferri, A.; Kang, Y.; Giannelis, E. P.; Elimelech, M.Superhydrophilic Thin-Film Composite Forward Osmosis Membranesfor Organic Fouling Control: Fouling Behavior and AntifoulingMechanisms. Environ. Sci. Technol. 2012, 46 (20), 11135−11144.

(81) Ling, M. M.; Chung, T. S.; Lu, X. M. Facile synthesis ofthermosensitive magnetic nanoparticles as “smart” draw solutes inforward osmosis. Chem. Commun. 2011, 47 (38), 10788−10790.(82) Li, D.; Zhang, X. Y.; Simon, G. P.; Wang, H. T. Forwardosmosis desalination using polymer hydrogels as a draw agent:Influence of draw agent, feed solution and membrane on processperformance. Water Res. 2013, 47 (1), 209−215.(83) Ge, Q. C.; Su, J. C.; Amy, G. L.; Chung, T. S. Exploration ofpolyelectrolytes as draw solutes in forward osmosis processes. WaterRes. 2012, 46 (4), 1318−1326.(84) Shaffer, D. L.; Yip, N. Y.; Gilron, J.; Elimelech, M. Seawaterdesalination for agriculture by integrated forward and reverse osmosis:Improved product water quality for potentially less energy. J. Membr.Sci. 2012, 415, 1−8.(85) Holloway, R. W.; Childress, A. E.; Dennett, K. E.; Cath, T. Y.Forward osmosis for concentration of anaerobic digester centrate.Water Res. 2007, 41 (17), 4005−4014.(86) Hoover, L. A.; Phillip, W. A.; Tiraferri, A.; Yip, N. Y.; Elimelech,M. Forward with Osmosis: Emerging Applications for GreaterSustainability. Environ. Sci. Technol. 2011, 45 (23), 9824−9830.(87) Hickenbottom, K. L.; Hancock, N. T.; Hutchings, N. R.;Appleton, E. W.; Beaudry, E. G.; Xu, P.; Cath, T. Y. Forward osmosistreatment of drilling mud and fracturing wastewater from oil and gasoperations. Desalination 2012, 312, 60−66.(88) McGinnis, R. L.; Elimelech, M. Global Challenges in Energy andWater Supply: The Promise of Engineered Osmosis. Environ. Sci.Technol. 2008, 42 (23), 8625−8629.(89) McCutcheon, J. R.; McGinnis, R. L.; Elimelech, M. Desalinationby ammonia-carbon dioxide forward osmosis: Influence of draw andfeed solution concentrations on process performance. J. Membr. Sci.2006, 278 (1−2), 114−123.(90) Elimelech, M.; Phillip, W. A. The Future of SeawaterDesalination: Energy, Technology, and the Environment. Science2011, 333 (6043), 712−717.(91) Kim, Y. C.; Park, S. J. Experimental Study of a 4040 Spiral-Wound Forward-Osmosis Membrane Module. Environ. Sci. Technol.2011, 45 (18), 7737−7745.(92) McGinnis, R. L.; Hancock, N. T.; Nowosielski-Slepowron, M.S.; McGurgan, G. D. Pilot demonstration of the NH3/CO2 forwardosmosis desalination process on high salinity brines. Desalination 2012,312, 67−74.(93) Yip, N. Y.; Tiraferri, A.; Phillip, W. A.; Schiffman, J. D.;Elimelech, M. High Performance Thin-Film Composite ForwardOsmosis Membrane. Environ. Sci. Technol. 2010, 44 (10), 3812−3818.(94) Zhao, S.; Zou, L.; Tang, C. Y.; Mulcahy, D. Recentdevelopments in forward osmosis: Opportunities and challenges. J.Membr. Sci. 2012, 396, 1−21.(95) Xiao, D.; Li, W.; Chou, S.; Wang, R.; Tang, C. Y. A modelinginvestigation on optimizing the design of forward osmosis hollow fibermodules. J. Membr. Sci. 2012, 392−393, 76−87.(96) McGinnis, R. L.; Elimelech, M. Energy requirements ofammonia-carbon dioxide forward osmosis desalination. Desalination2007, 207 (1−3), 370−382.(97) Ikeda, M.; Miyamoto, M. Forward Osmosis Device, andForward Osmosis Method. Japanese Patent JP2012091166, 2012.(98) Wilson, A. D.; Stewart, F.; Stone, M. L. Methods and Systemsfor Treating Liquids Using Switchable Solvents. United States Patent20130048561, 2013.(99) Stone, M. L.; Rae, C.; Stewart, F. F.; Wilson, A. D. Switchablepolarity solvents as draw solutes for forward osmosis. Desalination2013, 312, 124−129.(100) Koros, W. J. Evolving beyond the thermal age of separationprocesses: Membranes can lead the way. AIChE J. 2004, 50 (10),2326−2334.(101) Wigley, T. L. Coal to gas: the influence of methane leakage.Clim. Change 2011, 108 (3), 601−608.(102) Howarth, R. W.; Ingraffea, A.; Engelder, T. Natural gas: Shouldfracking stop? Nature 2011, 477 (7364), 271−275.



(103) Jenner, S.; Lamadrid, A. J. Shale gas vs. coal: Policyimplications from environmental impact comparisons of shale gas,conventional gas, and coal on air, water, and land in the United States.Energy Policy 2013, 53, 442−453.(104) Adams, M. B.; Edwards, P. J.; Ford, W. M.; Johnson, J. B.;Schuler, T. M.; Thomas-Van Gundy, M.; Wood, F. Effects ofdevelopment of a natural gas well and associated pipeline on thenatural and scientific resources of the Fernow Experimental Forest.United States Forest Service: Newtown Square, PA, 2011.(105) National Renewable Energy Laboratory. Federal EnergyManagement Program Maps; Golden, CO, 2013.(106) Blackwell, D.; Richards, M.; Frone, Z.; Batir, J.; Ruzo, A.;Dingwall, R.; Williams, M. Temperature at depth maps for theconterminous US and geothermal resource estimates. GeothermalResources Council Transactions; Geothermal Resources Council: Davis,CA, 2011; Vol. 35; available from http://geothermal.org.(107) Tester, J. W.; Anderson, B.; Batchelor, A.; Blackwell, D.;DiPippo, R.; Drake, E.; Garnish, J.; Livesay, B.; Moore, M.; Nichols, K.The future of geothermal energy; Massachusetts Institute of Technology:Cambridge, MA, 2006.(108) Driscoll, F. G. Groundwater and Wells, 2nd ed.; JohnsonDivision: St. Paul, MN, 1986.(109) United States Energy Information Administration. Maps:Exploration, Resources, Reserves, and Production. http://www.eia.gov/pub/oil_gas/natural_gas/analysis_publications/maps/maps.htm(accessed March 21, 2013).



http://geothermal.org

http://www.eia.gov/pub/oil_gas/natural_gas/analysis_publications/maps/maps.htm

http://www.eia.gov/pub/oil_gas/natural_gas/analysis_publications/maps/maps.htm

Documents

Desalination and Reuse of High-Salinity Shale Gas Produced